gpu - High Temperature when stress testing Intel Core i7 3770K
2014-07
I have bought an i7 3770K. I have tried to stress it, N
ear full load computation using CUDA I was surprised that it gets 99°C at 68% of the CPU load, is this normal? I have to change the cooling system?
I use a Cooler Master Hyper TX3 cooling (that is for i5 maybe). Should I use the normal fan sold with the CPU?
On Idle my CPU too is in the gap 32°-38° C the problem is under stress (only 68%) and I'm not overclocking the system!
The problem could be due to the GPU? But I have checked and I don't think so.
You will probably want to check a few things to see if this is causing the High CPU Temperatures.
- Make sure that you used Thermal Paste
- Don't use too much. A drop or two is sufficient. A paper thin spread is ideal.
- Clean any excess paste around CPU and Heatsink with Qtip and Denatured Alcohol. If you have paste that is overflowing, the heat cannot dissipate properly.
- Make sure that you have a solid connection with the heatsink.
- Make sure that the heatsink is receiving proper ventilation. (free of dust bunnies)
I have the Coolermaster N520 on my Ivy Bridge Core i7 and idle it will run about 32c.
I needed to do a small video compression today and took a snapshot of the CPU load and CPU temperature. With a large CPU load, it still was about 35c-37c.
I have a brand new laptop with an i7 2630QM CPU. Core #3 (Out of 1-4) always has a temperature 3-8C cooler than the other cores under normal operation. If I put 100% load on the core then it'll go up to the same temperatures as the other cores. Loading a different CPU core results in a 10C difference.
It seems to me that the temperature sensor for that core is off. Considering how small the CPU is, I seriously doubt that core #3 could have better cooling.
Should I be worried about the fact that one of the temperature sensors is inaccurate? (assuming it is inaccurate?)
The difference isn't much of a concern. But then again, you never posted the actual temperatures. The actual temperatures make more of a difference than the temperature differences.
My guess is die defects. It's common to have cores at different temperatures as long as they pass the stability test.
I wouldn't concern myself with it. Life is too short to concern yourself with arbitrary temperature differences.
Do you mean Core #3 as in 0,1,2,3 that three? If so then this may explain it:
If you look at a the die map of a Sandy Bridge CPU you will see that the graphics hardware is right next to core #3, with #0 all the way to the right.
If you mean 3 out of 1,2,3,4 then I have no idea.